How to Think about Microsoft’s New AI-Powered Knowledge Worker


How is Microsoft positioning artificial intelligence (AI)? Here is an excerpt from the Microsoft blog in September 2023.

“The era of AI is here, and it’s changing the way we do business. Technology is at an inflection point, offering organizations incredible opportunities to enhance human ingenuity with AI. This is transforming organizations and shaping the future of every industry, unlocking innovation and driving business value.”

To completely dismiss the enthusiasm and skies-are-the-limit thinking we’re hearing from every corner of the IT provider world would render me a “flat earther.” Instead, let’s consider Microsoft’s predictions in the context of the real world and see how AI will impact the Microsoft cloud business, namely Microsoft 365.

Teams and Apps for Enterprises are the two services built into Microsoft 365 Enterprise Suites that engage users and support collaboration for creating, sharing and managing documents. Recent Microsoft announcements indicate a very purposeful plan to embed AI-powered services, such as ChatGPT and other Microsoft-developed offerings, into Teams and Apps for Enterprises that could materially change how users engage and leverage these tools.

Microsoft uses the verb “amplify” to describe how AI services in Teams and Apps for Enterprises will reshape the future of work for knowledge workers. We’ve heard these marketing prognostications for over a decade now as Microsoft pioneered what they call the modern workplace, a specific strategy designed to grow the footprint for Teams and Apps for Enterprises across all its customers and tied it to their vision of empowering users to achieve more.

How Microsoft Is Offering AI to Enterprises

Microsoft is now advancing what it calls its second wave of AI-powered transformation, in which it brings AI to businesses in four different approaches. The approach of “AI as copilot” is the one we will focus on here. Microsoft 365 Copilot is a new “add-on” service to existing Microsoft 365 Enterprise subscriptions, providing users a virtual assistant that could eventually take over the creation of text, spreadsheet and presentation documents.

Many of us have read – or even written – documents that involved the help of ChatGPT. If you’ve used ChatGPT to write something, did you feel it was truly your work product in the end, produced with your unique style? Perhaps it was just something good enough to replace what you would have otherwise produced on your own, but still not yours.

What does embedded AI mean for today’s knowledge worker? Microsoft will certainly need to think about that question as it advances its Microsoft 365 ambitions and seeks to grow the volume and complexity of the user subscriptions it brings to market. Marketing around the “user” features of future subscriptions becomes increasingly moot if embedded AI is going to make the composition choices on behalf of knowledge workers.

What Are Possible Implications of AI for Knowledge Workers?

Knowledge workers today are routinely provisioned personal computers (PC) to support the tasks they process every day. With most of the AI compute resources residing in large datacenters, how much of the local compute provided by a PC will actually be needed? If the role of PC as we know it today is replaced by hyperscale datacenters, how will the bundling of the software services designed for knowledge workers change and what are the future cost implications?

While much of Microsoft’s marketing revolves around boosting personal productivity, the idea that knowledge workers will increasingly depend on and be impacted by AI warrants a more thoughtful conversation. Future software research and development will likely shift from providing more user-directed features to instead refining and enhancing the way knowledge workers ask questions of generative AI tools. It must also ensure that knowledge workers who generate ChatGPT output understand and can explain with confidence what has been produced and how it has been sourced. This requires more than simply selling subscriptions and handing over new tools to successfully transition to this new world.

As Microsoft moves forward with its AI agenda, the paradigm and design of today’s Microsoft 365 user services are going to change. What is becoming very clear is the enormous infrastructure cost associated with providing and processing AI services and workloads. The AI infrastructure investments required from Microsoft will most certainly be passed along to enterprise buyers in the form of subscription cost increases, user training requirements, new business risks and a higher level of complexity in doing business with Microsoft.

A key point here is that Microsoft has engineered software products for decades to be run on PCs.  As the transition to AI-embedded services within these software products becomes more widespread, PCs – and by extension the products designed to run on them – become less relevant as those services shift to Microsoft datacenters instead. This will bring forward the concept of Desktop as a Service (DaaS) and will fundamentally change the role of traditional PC in the AI-copilot world, as new requirements for compute capacity and governance become essential to managing the potential risk and unintended consequences of reimaging the enterprise knowledge worker.

How to Build an AI Control Plane

The Microsoft Copilot is just one of a myriad of tools and solutions that will continue to crowd the market and force leaders to continuously reevaluate the best tools and services - and the right way to buy and deploy them. To be successful with AI design, acquisition and deployment, enterprises must develop a perspective that grounds them in three essential questions:

  1. How do we secure our use of AI?
  2. How do we develop processes to ensure the responsible use of AI?
  3. How do we design and deploy self-financing AI?

In times of constant change, it's helpful to establish an AI control plane that will provide a consistent approach for adopting and maintaining new technology. Working with a qualified advisor to help establish these guiding principles can help you understand how doing business with providers like Microsoft is changing, how to adapt and how to avoid unexpected surprises and costs.

ISG AI Control Plane
Figure 1 ISG's AI Control Plane

Additionally, it's important to remember that, while AI is a relatively new field, qualified research can help support sourcing choices. ISG Provider Lens publishes studies to help enterprises navigate the Microsoft ecosystem and our professional advisors are ready to help you with your AI journey. Contact us to find out how we can help.